Some Random Wild Ideas

geniusMy list of hacking and business ideas has achieved critical mass. Meaning that I now have more than 100 items in my “someday” list.

So I decided to publish *some* of them. Those are the ones that I no longer have patience or coding skills for and those that I cant come up with a way to monetize (aka make money from)

Without further ado and in no particular order:

  • someone should launch a Bitcoin credit card. Maybe use the smartphone to do it. A card would be easier for physical use in stores and stuff. The banks might already be looking at this, so it would a tough job
  • a picture request site would be fun. You’d go to the site and request a picture of something or somewhere and someone in the network would take it. Hard to monetize I think. Plus there might be a lot of fake pictures taken (you know, search Google for something like the request).
  • an implementation of TCP/IP over Twitter would be hilarious
  • an implementation of a Pub/Sub queue (or other types of queues) would also be funny
  • someone should hack a converter to translate Erlang into Javascript. Probably using Emscripten, I dont know
  • implement the possibiliy of sending encrypted tweets. Probably through a FF/Chrome extension. People would publish their public key in their profile. The extension would be able to send an encrypted tweet to someone by reading their public key from the profile. The recipient of the tweet would be able to decrypt it using his private key (stored in some way or other, I dont know; local storage maybe)
  • a distributed search engine using browser Javascript webworkers to compete with Google. How? An extension for FF/Chrome would be needed. The extension would track and index every page the user visited. The results would be stored locally but also distributed to the other people in the network (ie using the extension)  using some kind of DHT. Users of the extension would then be able to make a query/search which would be retrieved through the DHT network. Privacy would be a problem
  • an Saas service that would provide people with an easy way to sell subdomains. Say that I have the domain. I would be able to sell subdomains (ie,, etc). The ecommerce part is pretty easy. The DNS magic needed would be the problem. And maybe not needed due to the new TLDs.
  • Portuguese government sites are a piece of crap. Someone should crawl the whole domain (and a few others) and provide a single interface to find information
  • the amount of crap that is tweeted everyday is astonishing. Mental farts are easy. Someone should build a better twitter (or better whatsapp, viber, etc) by creating a service that would only allow for one “tweet” per day. Make it count. I actually have a name for it and the domain:
  •  like I mentioned somewhere else in my blog (dont have the time right now to search for it), it would be possible to compete with twitter (or any other messaging service) by creating a distributed messaging system. Once again using browser extensions and DHTs. You’d have to some site to get the “app”. Whenever you “tweeted” (for want of a better word) the “tweet” would be stored in local storage but would also distributed to other users of the service in an encrypted form and stored in someone of those using a DHT scheme. Users would be able to request the “tweets” from the people they followed by querying the DHT network.
  • create a short URL database. Crawl the web for short urls (,,,, etc). Or just brute force them. And then find out and store the destination url. One would then be able to find out what short urls were available for each original url.  It would also make an excelent basis for a search engine. Sort of delicious but using actually important urls, those that people actually found interesting
  • register Create a centralized place for people to identify themselves. I guess already invalidated this idea
  • and, of course, a real full stack web development framework fully in Javascript. DB written in JS; app server written in JS; message queue in JS; template system using JS, browserside and just using HMTL. Prototype oriented (instead of the “class based OO insanity). With automatic object persistence (aka no ORM, the “vietnam of programming”). With inheritance via acquisition. For more information look at previous posts of mine and look at Zope.

Thats it. Have fun. I’ll only request 1% of shares if you make it big.


  1. Does this fit in any of your ideas?

  2. You can scratch one from the list: “crawl the whole domain”. Just Google with the “” option.

    Furthermore, definitely no money will come from there and no coding skills required :)